* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Fredericks transitions induced by light fields
Coherence (physics) wikipedia , lookup
Psychophysics wikipedia , lookup
Density of states wikipedia , lookup
Partial differential equation wikipedia , lookup
History of optics wikipedia , lookup
Derivation of the Navier–Stokes equations wikipedia , lookup
Nordström's theory of gravitation wikipedia , lookup
Circular dichroism wikipedia , lookup
Time in physics wikipedia , lookup
Photon polarization wikipedia , lookup
Equation of state wikipedia , lookup
Diffraction wikipedia , lookup
Wave–particle duality wikipedia , lookup
Wave packet wikipedia , lookup
Thomas Young (scientist) wikipedia , lookup
Theoretical and experimental justification for the Schrödinger equation wikipedia , lookup
Fredericks transitions induced by light fields B. Ya. Zel'dovich, N. V. Tabiryan, and Yu. S. Chilingaryan Erewn State University (Submitted2 December 1980) Zh. Eksp. Teor. Fiz. 8 1 , 7 2 4 3 (July 1981) The reorientation of the director of a nematic liquid cwstal induced by the field of a light wave is considered. An oblique (with respect to the director) extraordinary wave of low intensity yields the predicted and previously observed giant optical nonlinearity in a nematic liquid crystal. For normal incidence of the light wave on the cuvette with a homeotropic orientation of the nematic liquid crystal, the reorientation appears only at light intensities above a certain threshold, and the process itself is similar to the Fredericks transition. The spatial distribution of the director direction is calculated for intensities above and below threshold. Hysteresis of the Fredericks transition in a light field, which has no analog in the case of static fields, is predicted. PACS numbers: 61.30.Gd, 6 4 . 7 0 . E ~ 1. INTRODUCTION The nonlinear optics of liquid crystals has recently received a great deal of attention (see the review1). I n addition to the quite interesting quadratic nonlinearities (generation of the second harmonic2), recently there have been prediction^^-^ and experimental disc o v e r i e of ~ ~giant ~ ~ cubic optical nonlinearities of liquid crystals (nematic, cholesteric, and smectic), which a r e due to the reorientation of the director by light . ~ ~confields. Succeeding experimental s t ~ d i e s ' ~have firmed the presence of giant optical nonlinearities for a number of specific nematic liquid crystals (NLC) and experimental geometries. In particular, a nonlinear interaction was found" between a normally incident light wave and a homeotropically oriented cell of the liquid crystal (OCBP),and had a characteristic threshold dependence on the intensity of the l a s e r beam. This effect was explained qualitatively in Ref. 11 on the basis of an analogy with the Fredericks transition in the field of a light wave. In the present study we construct a quantitative theory of the Fredricks transition in the field of a light wave. In contrast to the simplest model of the Fredericks transition in static fields (see Refs. 12 and 13), here we take into account the following two facts: 1) the amplitude for the deformation of the director above threshold must be found taking into account the distortion of the longitudinal profile of the light wave itself a s it propagates in an inhomogeneous anisotropic medium; 2) if the transverse dimension of the beam is significantly smaller than the cuvette thickness, the threshold intensity itself depends on the transverse distribution of the intensity in the beam. In addition, the theory predicts that for certain liquid crystals the Fredericks transition is accompanied by hysterisis. 2. THE SYSTEM OF BASIC EQUATIONS We shall take the free energy per unit volume of an NLC i n the presence of a light field of complex amplitude E to be of the form F [ergicm'l ='/2Kl, ( d i n)2+112K?2(n ~ rot n)' L'11K13[n rotnIZ-t.(nE) (nE')/lGrr-~_IE 1 2 / 1 6 ~ . 32 Sov. Phys. JETP 5 4 1 ), July 1981 Here K,, a r e the Frank constants, n i s a unit vector in the direction of the director (nand -n will be assumed equivalent), and the permittivity tensor of an NLC a t the frequency of the light field w is &,*=&,6,*+(E,,-E,.) n,n~. In addition, we have used in (1) the notation c , = E,,- cl. The complex amplitude of a quasimonochromatic field E ( r , t) is determined in a self-consistent manner from the solution of Maxwell's equations with E,r (r, t) =8~6t~+&cnt(r, t) t). However, we emphasize that to obtain the variational equations for the director n(r, t) it is necessary to a s sume that the amplitude of the electric field E ( r , t) is fixed. We shall take the density of the dissipative function R (in erg/cm3 s e c ) to be (2) R='/,I~(nn). T h e variational equations for n(r ,t) have the form --6F 6ni 6R a 6F =Sn,-ax, s(anilaxA) 6ni ' (3) is an undetermined Lagrange multiplier that where ensures that the condition I nl = 1 is satisfied. 3. THE FREDERICKS TRANSITION IN BROAD LIGHT BEAMS Let us consider a homeotropically oriented NLC cell occupying a layer 0 a z c L. We shall assume that the light field has nonzero components Ex and E,, while E , = 0. The unperturbed direction of the director in n o =e,. We shall also assume that the perturbation of the director does not move it out of the xz plane. Then n(r) =e, cos q+e. sin cp, where cp =cp(r). If the transverse dimensions of the beam a, in the xy plane a r e much larger than the cuvette thickness L and the beam itself is almost a plane wave, the field components Ex and E, can be assumed to depend only on z ; more precisely, E=[e,E,(z) +e,E, (z) ] esp ( i o s x l c + i k , z ) . (4) Here s =sins and ff is the angle of incidence of the O 1982 American Institute of Physics 32 cp(z, nkz t)=C sin-exp L rht, cp, I F r o m (7) we find that f o r FIG. 1. Unperturbed director vector no perpendicular to the cell walls. The wave vector k'of the light field makes an angle with the normal to the cell walls. light wave on t h e cell in a i r ( s e e Fig. 1). Under these conditions the slope of the d i r e c t o r rp a l s o depends only on 2 . Now the variational equations (3) take the f o r m ascp ( a , ,sinzcp+K,, cos2cp)- a z2 -(K.-K,,)sin cp cos cp acp (x) Let u s f i r s t give a qualitative picture of the F r e d e r i c k s transition. When a plane light wave falls a t normal incidence on a homeotropic cell the e l e c t r i c field of the wave is exactly perpendicular to the director. A t a positive value of E, it would be energetically favorable t o orient the d i r e c t o r in the direction of the field. However, this is prevented by the homeotropic orientation of the d i r e c t o r by the curvette walls. F u r t h e r m o r e , in the f i r s t approximation in the light intensity t h e orientational effect of the field on the unperturbed d i r e c t o r is absent, in other words, for E =e,Ex the function cp(z) = 0 is a n exact solution to equation (5). Above a certain threshold value of the intensity the solution (p(z) = 0 will no longer be stable. T o determine this threshold it is convenient to consider the linearized [that i s , a t s m a l l (p(z)] equation (5). H e r e we m u s t take into account the fact that f r o m the equation d i v D = O we have T h e r e f o r e , to t e r m s linear in (p and equation (5) takes the f o r m a perturbation of the f o r m s i n ( n z / l ) begins to grow exponentially. sin(nz/~) T h e steady-state value of theamplitude cp =, , ,pc in the r e g i m e above threshold m u s t now be xetermined with allowance for the nonlinear t e r m s in equation (5). I t is v e r y important that in this approximation we m u s t include a solution consistent with (p(z) to Maxwell's equations in a n inhomogeneous anisotropic medium. F o r a wave of the f o r m (4) these solutions c a n be obtained in the geometrical optics approximation ( s e e Refs. 14 and 15): E,(z) = A ( E , , - s ~ ) ' ~ ~ ' ~ ( ~ ' , E.. E,(z)=-A (e.,--sa)"+s(eIle,)'" e,, (E,,-s')" e i.rcz,. T h e advance of the optical phase $ ( z ) is equal t o o *(:)=-J - " - s e . , ( ~ ' ) + I e , ~ ~ ~ ( e . , ( z ' ) -I"'s ~ )d Z , e,.(z') 0 A s a l r e a d y noted, s =sins, where a is the angle of incidence in a i r , and t h e E,, a r e s,,=e, sin cp ( z ) cos cp ( z ) . E,,=E,+E. cosZq ( z ) , (9) T h e constant A c a n b e e x p r e s s e d in t e r m s of the power flux density; namely the z component of the Poynting vector is P , = C ( E ~ ~ lz/8n ~ ~ ) ' ~ ~ A and is independent of 2. I n the c a s e of exactly n o r m a l incidence (s =0) equation (5), taking into account (8) and (9), has the followwe have E , = - E , ( ~ E ~ / E ~ , ing f o r m i n the stationary c a s e : d Z 9-(K,,-K,,)sin cp cos cp ( K i tsinzcp+K~scoszcp)dz2 e.elIe,lAI2 sincpcos cp 8%(e,+e. cosZcp)" = O. + Equation (6) is analogous to the linearized equation for the behavior of the d i r e c t o r n e a r the F r e d e r i c k s transition in a s t a t i c field E ,,,. = e x E,,, (cf. Ref. 13, section 4.2). T h e r e a r e two differences, however. of the s t a t i c c a s e in (6) we F i r s t , instead of the have I E, 12/2-the mean s q u a r e of the amplitude of the light field. Secondly, in a s t a t i c field E,,,the condition d i v D = O has no effect on the field vector E when the dir e c t o r is inclined. In c o n t r a s t t o this, in the problem with a light field, a component E, a p p e a r s because of the condition div D = 0. e,,,, As a r e s u l t , in equation (6) we have a n additional factor the r o l e of which reduces to a c e r t a i n inc r e a s e of the threshold. Expanding (p(z, t ) in a F o u r i e r s e r i e s and taking the boundary conditions ~ ( =z0 , t ) = cp(z = L, t ) = 0 into account, we obtain f r o m (6) 33 Sov. Phys. JETP 54(1), July 1981 T h e e x a c t solution of this equation will be given in Sec. 5 below. H e r e we s h a l l r e s t r i c t ourselves to only taking into account t e r m s of o r d e r cp and (p3 in this equation, a procedure valid n e a r the F r e d e r i c k s threshold. T h e s e a r c h f o r a solution of the f o r m c p " f r sin ( n z i L )+cp, sin ( 3 n z l L )t .. ., which automatically s a t i s f i e s the boundary conditions, leads to the following expression for (p,: 9 e. cpt2=2 ( 1 ---4 ell l,! ) -' P-P ,, -7q-l - H e r e (p, rp:. T h e r e f o r e rp, is proportional to the s q u a r e r o o t of t h e e x c e s s of the intensity P above threshold P, . . Zel'dovich eta/. 33 The advance of the optical phase #(L) over the cuvette thickness L can be obtained from (8b); expanding (8b) to terms -d we obtain -' P-P F, C 4. THE FREDERICKS TRANSITION I N NARROW BEAMS In the case where the transverse dimension of the beam a, is smaller than o r on the order of the cuvette thickness L, the Frank energy due to the transverse gradients of the director becomes dominant. Let us first estimate the order of magnitude of the threshold power of the Fredericks transition. The energy of the perturbed state is Here 6 2 , ~is the volume occupied by the disturbance and K is the Frank constant. This expression gives a stable state cp, = O a t small IE 12, and instability s e t s in a t As seen from expression (14), IE l2a; = const a t a,<< L. In order to exactly determine the threshold of the Fredericks transition, it is necessary to solve the three-dimensional problem of the stability of the solution cp =O. We shall consider this problem for several particular cases. a. The single-constant approximation. In the singleconstant approximation we can assume even for a narrow beam that the vectors E and n lie in the xz plane. ) have Then in the approximation linearized in ~ ( r we Here P , ; is determined by formula (12) using the fact that K l l =K. With the substitution cp(r, t) = g'(r) e x p r t , equation (15) becomes of q for a "potential" of the form (18) will be1' The eigenvalues of P from (17), corresponding to the ) is P = n 2 / ~ ' . The perturbaeigenfunction ~ ( z=sin(nz/l) tion of the director field cp'(r) will grow exponentially in time at r = q - ' ~ ( q-P) > 0. Then, defining the threshold intensity of the Fredericks transition a s that value of Poa t which a bound state first appears in the potential well U a l/cosh2 px, we obtain - P th (l+bL/n). (20) The quantity BL can be viewed a s the ratio of the cuvette thickness L to the beam diameter d: PL- ~ / d . b. Another case which can be solved analytically and by which we can approximate the r e a l distribution of the intensity in the beam of a light wave corresponds to a function P(r) of the form const for p G p ~ (2 1) where p is the distance from the beam axis in the transverse direction. Writing equation (16) in cylindrical coordinates and making the substitution cp'(r) = L(P)X(Z), we find The solution of (22a) is a zeroth-order Bessel function. Let P ( p C p o )> ( L l n ) ? h P ~=PF, , [in expression (23) we have s e t h =$/L', since directorfield perturbations which increase with time exist only for r = q - l ~ ( h- g ) > 0, where g, just a s P in case (a), equals I?/L']. I t is easy to s e e that the condition (23) is necessary for the existence of a nontrivial, that is, not identically equal to zero, solution to the system of Eqs. (22), compatible with the boundary conditions. Then the solution of Eq. (22a) will be S(P)= Let us first consider the case where a "ribbon" light beam falls on the medium, P(r,) =P(x). Then after separating the variables in equation (16) we find by means of the substitution cpl(r) =E(x)~(z) (23) (2 4) for p>p0 where cl and c2 a r e constants and J, and K, a r e Bessel functions. The condition of continuity of the logarithmic derivative at the point p = po gives Equation (17a) is the one-dimensional Schrijdinger equation for the potential U - -P(x). It is possible to solve this equation analytically for a relation, for example, of the form P ( z )= P ~ I C ~ ~ ~ ~ Z , (18) that is, for a function which gives a good qualitative approximation of a Gaussian (Fig. 2). The eigenvalues 34 Sov. Phys. JETP 54(1), July 1981 FIG. 2. Form of the function 1/cosh2x. The dashed line shows the function exp(-x2). Zel'dovich et a/. 34 =K, (25) (p) / KO(y). In the case n p , / ~ < <1, if in addition we assume that 3(LL <I, (26) PF~ we have from Eq. (25) This occurs, for example, for the nematic crystal PAA, for which &,= 0.9; K,, =4.5 x9.5 x lom7dyne [at the temperature ~ = 1 2 5 " C(Ref. 13)] and B=-0.03 (B =0.7 for MBBA and for B=0.06 for OCBP). T o study the Fredericks transition for B s 0 we integrate Eq. (10) with respect t o z after multiplying it by 2dq/dz. After determining the integration constant in t e r m s of the maximum angle of deviation of the director from the unperturbed direction cp, we obtain P th = P F[I-2LZlilZp.' ~ In (np,/L) 1. (27) As s e e n from expression (27), the condition (26) is satisfied for ln(np0/L)>> -2, and this strengthens considerably the inequality np0/L<< 1. In the case where p o 2 L, we obtain from Eq. (25) JO,LK,(~P~~L) npoK,(npdL) . 1'1 dq (di) - e,' P ~I-(e./e,l)sin~ql"'-[I-(~./ell)sin2q,l'~ rX.. ( I - ) sin'q) [I- (e./~,~)sin~q]"'[ 1- ( ~ ~ / e ~ ~ ) s i n ~' ( o , , , " ~ (32) Taking into account the fact that the maximum angle of deviation cp, is reached a t the center of the cell, that is, for z = ~ / 2 , we get from (32) (28) Ym (1-k sinZq ) [I-(e./e,,)sin'q]'~[l- (en/~,)sin2 cp,]" )" dq Here Jo,= 2.4 is the f i r s t z e r o of the Bessel function ~ ~ ( 2 In ) . the limit po>>L formula (28) gives PI, =P,,, in agreement with the result obtained in Sec. 3 for broad beams. j( In the case where the beam radius is of the order of the cuvette thickness, np,/L=: 1, Eq. (25) can be solved graphically. A s a res~!lt we find Assuming that the power density P of the radition falling on a cell of the NLC is close to the threshold value for the Fredericks transition, that i s , rp,c 1 , let The us compute the integral in (33) up to t e r m s -rp.: solution of the resulting biquadratic equation for cp, has the form Therefore, in all the cases studied we have PI, -PI, 1/L2, whereas the coefficient of proportionality is determined by the ratio p , / ~ of the beam radius to the cuvette thickness. - In the case of more than a single constant the threshold of the Fredericks transition can be determined analytically for a ribbon beam of the form P(r,) =P(y)polarized along the x axis. The equation linearized in ~ ( r in) this case has the form Introducing the new variable y' =(K,,/K~,)"'~, Eq. (30) can be written in the form (17) by replacing x by y and using the definition of P, from (12). After using the solution to (17), we obtain for a light wave of the form P(y) = Po/cosh2,9y [ 1- (E,,/E,\) si$ q ~ ' ~ - [ f ~- e ~ / e l l ) sq,,,]'h ina (33) where For the liquid crystals MBBA, OCBP, and PAA the values of G a r e almost identical a t G;. 0.06. Before discussing formula (341, we note that it can in principle be obtained from the condition of the minimum of the free energy after expanding it in a s e r i e s in p, up to t e r m s Apart from a coefficient that renders it dimensionless, the free energy has the form At the point determined by'formula (34) we have Therefore, when the beam dimensions a r e slightly smaller than the curvette thickness the inclusion of more than one constant leads to small corrections. We can expect this to hold for other beam shapes, also. 5. HYSTERESIS OF THE FREDERICKS TRANSITIONS I N BROAD BEAMS Formula (11) becomes invalid for the maximum angle of deviation of the director from the unperturbed direction if the parameters of the liquid-crystal medium a r e such that @& IPm FIG. 3. Dependence of the free energy iP on cp, for different values of the light field power P: I-P 4 Pa,2 - P = P & , &Pth . 4 35 Sov. Phys. JETP 54(1), July 1981 *P < Prr, -P > g,. Zel'dovich eta/. 35 the field of the director in this case were discussed in Secs. 3-5. Here we shall study the nature of the reorientation of the director by the field of an extraordinary light wave in the intermediate region, that is, when the wave propagates a t small angles to the director. FIG. 4. Hysteresis of the Fredericks transition. The arrows indicate the direction of variation of the power of the light field P. Substituting the expressions for the light fields (8a) into Eq. (5) and including terms up to third order in the small quantity rp = rp,sin(nz/~) (that is, assuming that the power of the wave is less than o r of order P, :) and of first order in s =sins, we find Therefore, we should use the plus sign in formula (34) (the condition that the free energy be a minimum). - If BaO the quantity cp, will be r e a l a t C = 1 ( P / P , , ) ' ~ S O , that is, P,, =P,,. If B< 0, then in order that rp, be r e a l it is sufficient to require that the expression under the square root in (34) be positive, B2- 4GC 3 0 , from which we have where sin y = &-In s i n a and y is the angle of refraction of the wave in the cell of the NLC. In the case of weak fields P<< P,, and large anisotropy of the permittivity, Eq. (39) duplicates the results obtained earlier4: cp,>,=e.IE IZLZ sin y/2n4K,,. However, we s e e from (36) and (37) that when the power density of the incident radiation is P =PI,, the quantity cp, given by (34) is a point of inflection for the free energy function (see Fig. 3). As seen from Fig. 3, for values PI,< P < P,, the function 9(cp,) has a local minimum at cp,+ 0, where The point p, = O becomes unstable only a t P>P,,. From the above discussion it is clear that a s the power density of the light wave increases from z e r o the Frede r i c k transition occurs at P=P,,, s o that a free-energy minimum corresponding to cp#, 0 and occurring at PI, < P < P,, is unattainable because of the presence of the potential barrier. For the inverse transition, that is, a s P decreases from the region of larger P, ., the freeenergy minimum reached a t P < P,, and corresponding to p,# 0, while not an absolute minimum of the function 9 [as seen from (3811, is separated from it by the potential barrier. The barrier only disappears a t P=P,, and then we have rp, = 0 (Fig. 4). 6. PROPAGATION OF A LIGHT WAVE AT SMALL ANGLES TO THE DIRECTOR As already noted, an extraordinary wave propagating a t an angle to the director induces strong nonlinear optical effects a t very low powers P C P , , (Refs. 3-9). When a light wave is normally incident on a cuvette with homeotropic orientation of the NLC , however, reorientation of the director is possible only above some threshold value of the power of the wave. The distortions of Since the formulas a r e quite awkward we shall not give the analytic form of the solution of Eq. (39). The graph of the function c p , ( ~ / ~,), is shown in Fig. 5 for the parameters of the liquid crystal OCBP and for different angles y. The behavior of the cell in the oblique-incidence case discussed here is analogous to the case of a cell in an oblique magnetic field-in both cases there is no rigorous F r e d e r i c k transition (cf. Ref. 13, $4.2.3). 7. DISCUSSION Let us estimate numerically the value of the threshold power for the Fredericks transition for the nematic crystal OCBP used in Ref. 11. The parameters of the liquid crystal a r e & f a = 1.66, &i0= 1.53, Kll = 4 x dyne, the curvette thickness is dyne, K,, = 7 x L = 150 p m , and the beam radius is p, = 50 pm. Substituting the Frank constant K =0.5 (K,, +K,,) into the expression for P I,, we find from formula (29) This value is in very good agreement with the experimental value" For the cuvette thickness L =50 pm and the same values of all the other parameters the value of the threshold power estimated from (28) is This value is 3.1 times larger than that for the Frederi c k transition in a cuvette of thickness L = 150 pm, which is also in very good agreement with the results of Ref. 11. FIG. 5. Dependence of the maximum angle of deviation cp, on P/P,, in the case of a light wave incident on a cell at small angles: 1-y = l o 0 , Z-y= 5". 36 Sov. Phys. JETP 54(1), July 1981 The temperature dependence of the threshold power is determined by the temperature dependence of the Frank constant and of the permittivity. Bearing in mind the fact that the Frank constant K varies with temperature a s the square of the order parameter K-5? while the anisotropy of the permittivity &,-S (Ref. 12), we can Zel'dovich e t a / . separate the temperature-dependent part in expression (12) for P,, in the form PF,a c,E,~&;"~. AS the temperature' increases, E, and d e c r e a s e s while E, increases; f r o m this we s e e that a s the temperature increases the threshold power decreases. Introducing the notation we can write P,, & ,, . Since &;I2 depends weakly on the temperature (for example, in the region of the MBBA mesophase &t12changes by only 0.01 of its value), a s the point of the phase transition to a n isotropic liquid is approached the threshold power decreases in proportion to c,, that is, to the o r d e r parameter S. Here we note the following. At s m a l l values of and k, a s is the c a s e near the critical point, Eq. (33) determiningthe maximum angle of deviation of the director cp, a s a function of the power of the light field can be simplified. For this we must expand the integrand in powers of and k. Then the integral on the lefthand side of Eq. (33) is expressed in t e r m s of elliptic integrals and the resulting equation implicitly determines the function cp, = cpm(P). In the limit where P/P, , << 1 (that i s , (pm=n/2 6,6<< 1) the function cpm(P)can be determined explicitly: - As was shown in Sec. 5, for certain types of liquid crystals the Fredericks transition can be accompanied by hysteresis. Let us calculate f o r the c a s e of the well studied liquid crystal PAA the main characteristics of this phenomenon. As s e e n from formula (37) with B =-0.03 and G =0.06, the powers a t which the Fredericks effect in PAA i s switched on and off differ little: P,, = 0.9925 T h e angle corresponding to the appearance of a local minimum of the free-energy function [that i s , pmcorresponding to the inflection point of the function @(cpm)]is c p , ~ ( - ~ / 2 ~ )=0.5. "~ ,. Let us demonstrate that thermal fluctuations of the direction of the director do not lead to surmounting of the potential b a r r i e r between the minima of the f r e e energy function. F o r this we note that thermal fluctuations of the direction of the director would cause the hysteresis t o disappear if the mean angle of the fluctuation deviations were on the order of o r larger than the difference of the angles corresponding to the local minimum and maximum of the f r e e energy. T h i s difference is easily calculated using formula (34) for some value of the power P such that P,,.< P < P,,(that i s , O< C < B'/ 40). For example, for C = B ' / ~ D we find A q =O.4, and since ((6cp)2)"2<<1 for thermal fluctuations of the direct o r the local minimum of the f r e e energy CP.(cp,> 0) is stable to thermal fluctuations. In concluding this discussion we note that a Frederi c k ~transition induced by light fields in a NLC with &,> 0 can occur only when the light wave is normally incident on a homeotropically oriented NLC cell. I t is impossible to have a Fredericks transition induced by 37 Sov. Phys. JETP 5411, July 1981 a n ordinary light wave normally incident on a planarly oriented NLC cell. T h i s can be understood when it is realized that the electric field of an ordinary light wave in a NLC remains perpendicular to the director a s it propagates through the medium. Strong nonlinear optical effects a r i s e if the polarization of the normally incident cell makes s o m e angle with the director. We have considered these effects in a n e a r l i e r study.' T h e authors a r e grateful to E. I. Kats f o r a discussion of these problems. Note added inwoof. We have been kindly informed by the authors of a paper submitted t o this journal that they have also studied the theory of light-induced F r e d e r i c k s transitions (A. S. Zolot'ko, V. F. Kitaeva, N. N. Sobolev, and A. P. Sukhorukov, Zh. Eksp. Teor. Fiz. 81, 933 (1981) [Sov. Phys. J E T P 54, No. 9 (1981)l) 's. 1vI. Arakelyan, G. A. Lyakhov, and Yu. S. Chilingaryan, Usp. F i z . Nauk 131, 3 (1980) Fov. Phys. Uspekhi 23, 245 (1980)l. 's. M. Arakelyan, G. L. Grigoryan, S. T s . Nersisyan, M. A. Nshanyan, and Yu. S. Chilingaryan, ZhETF Pis. Red. 28, 202 (1978) [ J E T P Lett. 28, 186 (1979)l. 3 ~ Ya. . Zel'dovich and N. V. Tabiryan, ZhETF Pis. Red. 30, 510 (1979) [ J E T P Lett. 30, 478 (1980)l. 4 ~ Ya. . Zel'dovich, N. F. ~ i l i p e t s k i r ,A. V. Sukhov, and N. V. Tabiryan, ZhETF Pis. Red. 31, 287 (1980)[JETP Lett. 31, 263 (1980)l. 'B. Ya. Zel'dovich and N. V. Tabiryan. Preprint F U N , No. 61, 1980. 6 ~ .Ya. Zel'dovich and N. V. Tabiryan, Preprint FIAN, No. 62, 1980. 7 ~ Ya. . Zel'dovich and N. V. Tabiryan , Preprint F U N , No. 63, 1980. 'B. Ya. Zel'dovich and N. V. Tabiryan, Kvant. Elektronika (~kIoscow)7 , 770 (1980) [Sov. J. Quantum Electronics 1 0 , 440 (1980)l. 9 ~ F. . Pilipetski, A. V. Sukhov, N. V. Tabiryan, and B. Ya, Zel'dovich, Optical Communications 37, 280 (1981). '9. B. Pakhalov, A. S. Tumasyan, and Yu. S. Chilingaryan, Abstracts of Reports a t the Tenth Int. Conf. on Coherent and Nonlinear Optics, Kiev, October 14-17, 1980, P a r t I, p. 18. "A. S. Zolot'ko, V. F. Kitaeva, N. Kroo. N. N. Sobolev, and L. Chilag, ZhETF Pis. Red. 32, 170 (1980)lJETP Lett 32, 158 (1981)l. "P. G. De Gennes, The Physics of Liquid Crystals, Oxford University P r e s s , 1974 (Russ. transl., Mir, 1977). 1 3 ~ M. . Blinov, Elektro- i magnitooptika zhidkikh kristallov (Electro- and Magneto-optics of Liquid Crystals), lvbscow, Nauka, 1978. 1 4 ~ n a l i t i c h e s k irnetody e v teorii rasprostraneniya i difraktsii) voln (Analytic Methods in the Theory of Wave Propagation and Diffraction), ed. S. V. Butakovaya, Moscow, Sov. radio, 1970. '%. Ya. Zel'dovich and N. V. Tabiryan, Zh. Eksp. Teor. Fiz. 79, 2388 (1980)[Sov. Phys. J E T P 52, 1210 (1980)l. 1 6 ~ .D. Landau and E. M. Lifshitz, Kvantovaya mekhanika, Moscow, Nauka, 1974 (Engl. transl. Quantum Mechanics: Nonrelativistic Theory, New York, Pergamon, 1977). Translated by Patricia Millard Zel'dovich et at.